Geoff Cumming has written a great article in The Conversation about problems with using p-values for determining importance when using statistics. He recommends focusing on measuring effect sizes by using confidence intervals. It is well worth reading, but you should also definitely watch his YouTube video about the Dance of the P-values.
His basic argument is that confidence intervals represent the information in an experiment much more reliably than p-values. P-values are variable between possible replicates of an experiment and do not indicate how variable they might be; the same basic experiment can give p-values that range from <0.001 to 0.5, depending on how the randomized data fall out. P-values dance around. In contrast, confidence intervals are less variable, and also indicate the magnitude of possible variation that might occur in a replicate of the experiment.
Geoff also has experimental evidence that using confidence intervals improves statistical interpretation. Geoff’s work is primarily in the area of psychology, but the same issues occur in other disciplines. Medicine moved to confidence intervals, with apparent improvements in the science and patient outcomes. Ecology and conservation science have plenty of scope for improvement.
Want to read more about null hypothesis significance testing? Check out this.